Regulation of plant growth by cytokinin
Tomas Werner*, Vaclav Motyka, Miroslav Strnad, and Thomas Schmulling*
*Centre for Plant Molecular Biology (ZMBP)Allgemeine Genetik, Universitat Tubingen, Auf der Morgenstelle 28, D-72076 Tubingen, Germany;
Institute of Experimental Botany, Academy of Sciences of the Czech Republic, Rozvojova 135, CZ-16502 Prague 6, Czech Republic; and
Palacky University and Institute of Experimental Botany ASCR, Department of Botany, Laboratory of Growth Regulators,
Slechtitelu 11, CZ-78371 Olomouc, Czech Republic
Communicated by Jozef S. Schell, Max Planck Institute for Plant Breeding Research, Cologne, Germany, June 15, 2001 (received for review
February 15, 2001)
Cytokinins are a class of plant-specific hormones that play a central
role during the cell cycle and influence numerous developmental
programs. Because of the lack of biosynthetic and signaling mu-
tants, the regulatory roles of cytokinins are not well understood.
We genetically engineered cytokinin oxidase expression in trans-
genic tobacco plants to reduce their endogenous cytokinin content.
Cytokinin-deficient plants developed stunted shoots with smaller
apical meristems. The plastochrone was prolonged, and leaf cell
production was only 3 4% that of wild type, indicating an absolute
requirement of cytokinins for leaf growth. In contrast, root meri-
stems of transgenic plants were enlarged and gave rise to faster
growing and more branched roots. These results suggest that
cytokinins are an important regulatory factor of plant meristem
activity and morphogenesis, with opposing roles in shoots and
roots.
C ytokinins were discovered during the 1950s because of their
ability to induce plant cell division (1). Shortly after their
discovery, Skoog and Miller coined the auxincytokinin hypoth-
esis of plant morphogenesis (2). The hypothesis predicted that
cytokinin, together with auxin, plays an essential role in plant
morphogenesis, having a profound influence on the formation of
roots and shoots and their relative growth.
Chemically, natural cytokinins are N6-substituted purine de-
rivatives. Isopentenyladenine (iP), zeatin (Z), and dihydrozeatin
(DZ) are the predominant cytokinins found in higher plants. The
free bases and their ribosides (iPR, ZR, DZR) are thought to be
the biologically active compounds. Glycosidic conjugates play a
role in cytokinin transport, protection from degradation, and
reversible and irreversible inactivation (3).
Numerous reports ascribe a stimulatory or inhibitory function
to cytokinins in different developmental processes such as root
growth and branching, control of apical dominance in the shoot,
chloroplast development, and leaf senescence (4). Conclusions
about the biological functions of cytokinins have mainly been
derived from studies on the consequences of exogenous cytoki-
nin application or endogenously enhanced cytokinin levels (5, 6).
Up to now, it has not been possible to address the reverse
question: what are the consequences for plant growth and
development if the endogenous cytokinin concentration is de-
creased? Plants with a reduced cytokinin content are expected
to yield more precise information about processes cytokinins
limit and, therefore, might regulate. Unlike other plant hor-
mones such as abscisic acid, gibberellins, and ethylene, no
cytokinin biosynthetic mutants have been isolated (7).
The catabolic enzyme cytokinin oxidase (CKX, ref. 8) plays
possibly the principal role in controlling cytokinin levels in plant
tissues. CKX activity has been found in a great number of higher
plants and in different plant tissues (8). The enzyme is a
FAD-containing oxidoreductase that catalyzes the degradation
of cytokinins bearing unsaturated isoprenoid side chains. The
free bases, iP and Z, and their respective ribosides are the
preferred substrates. The reaction products of iP catabolism are
adenine and the unsaturated aldehyde 3-methyl-2-butenal (8).
Recently, a cytokinin oxidase gene from Zea mays has been
isolated (9, 10). The manipulation of CKX gene expression could
www.pnas.orgcgidoi10.1073pnas.171304098
partially overcome the lack of cytokinin biosynthetic mutants
and might be used as a powerful tool to study the relevance of
iP- and Z-type cytokinins during the whole life cycle of higher
plants. In this article, we report the cloning of four putative CKX
genes from Arabidopsis thaliana and the results of their systemic
overexpression in transgenic tobacco plants. Our data indicate
an important role for cytokinins in plant growth regulation via
a differential influence on the number andor duration of cell
division cycles in the root and shoot meristems.
Materials and Methods
Gene Cloning. The genomic sequences of the AtCKX1, AtCKX2,
AtCKX3, and AtCKX4 genes were amplified by PCR from DNA
of A. thaliana accession Col-0. Oligonucleotide primers were
designed according to the published genomic sequences of
AtCKX genes [GenBank accession nos. AC002510 (AtCKX1),
AC005917 (AtCKX2), AB024035 (AtCKX3), and AL079344
(AtCKX4)] and had 5 and 3 overhangs with SalI or KpnI
restriction sites, which permitted subcloning in the vector
pUC19. The length of the amplified sequences were 2,235 bp
(AtCKX1), 3,104 bp (AtCKX2), 3,397 bp (AtCKX3), and 2,890 bp
(AtCKX4). Genes were sequenced and inserted into vector
pBINHygTx under the transcriptional control of a constitutively
expressed 35S promoter (11). The cDNA of AtCKX2 was cloned
by reverse transcriptionPCR from total RNA of AtCKX2
transgenic plant tissue with the OneStep reverse transcription
PCR kit (Qiagen, Chatsworth, CA). The PCR products were
sequenced and positioned under control of the GAL1 promoter
in the yeast expression vector pYES2. The control strain har-
bored only the empty vector. Induction of gene expression by
galactose was carried out for 6 h as suggested by Invitrogen.
Plant Transformation and Plant Culture. Nicotiana tabacum L. cv.
Samsun NN leaf explants were transformed and regenerated as
described (12). At least 15 independent transformants showing
very similar phenotypes were obtained for each of the four genes.
Plants were cultured in vitro on MS medium or in a glass house
with 15-h light9-h dark cycles, 20C during the dark period and
24C during the light period. Characterizations of the transgenic
tobacco were carried out on T2 progeny obtained by selfing.
Phenotypic changes noted for the independent transformants
were very similar and differed only gradually. Independent
transformants, confirmed by Northern blot analysis andor by
PLANT BIOLOGY
measuring the cytokinin oxidase activity, looked similar to the
transformants shown in Fig. 2B. Quantitative growth parameters
were obtained from at least ten individuals of two independent
clones (AtCKX1-28, AtCKX1-50 and AtCKX2-38, AtCKX2-40,
respectively). Segregation analyses of the hygromycin resistance
Abbreviations: CKX, cytokinin oxidase; iP, isopentenyladenine; Z, trans-zeatin; SAM, shoot
apical meristem.
Towhom reprint requests should be addressed. E-mail: thomas.schmuelling@zmbp.uni-
tuebingen.de.
The publication costs of this article were defrayed in part by page charge payment. This
article must therefore be hereby marked advertisement in accordance with 18 U.S.C.
1734 solely to indicate this fact.
PNAS August 28, 2001 vol. 98 no. 18 1048710492
gene indicated one or two (AtCKX2-40) chromosomal T-DNA
insertion loci. For the sake of clarity, the results for only one
clone are shown in Figs. 2 (C and D) and 3 (C and D).
RNA Preparation and Blot Analysis. Total RNA extraction from leaf
tissue and Northern blot analysis, with 50 g of total RNA, was
carried out essentially as described (13).
Histological Analysis. Plant tissue was fixed and embedded in LR
White (Plano, Wetzlar, Germany) according to ref. 14, and 2.5
M thin sections were stained with 0.1% toluidine blue. For
DNA staining, roots were fixed in ethanol:acetic acid (6:1),
incubated for 15 min in a solution of 4,6-diamidino-2-
phenylindole (DAPI, 1 gml H2O), and washed three times
with water. Scanning electron microscopy was carried out ac-
cording to ref. 15.
Quantitative Analysis of Cytokinin Oxidase Activity. The standard
assay for CKX activity was based on the conversion of [2-3H]iP
to adenine as described (16).
Quantitative Analysis of Cytokinin Content. Cytokinin extraction,
immunopurification, HPLC separation, and quantification by
ELISA methods were carried out as described (13).
Results
The Structure of the AtCKX Genes. A BLAST search of the GenBank
database identified six cytokinin oxidase-like genes in Arabidop-
sis that code for enzymes with 3241% amino acid identity to the
maize protein and 3366% amino acid identity between indi-
vidual family members. The gene structure is partially conserved
between maize and Arabidopsis. The predicted Arabidopsis genes
have five exons and four introns, whereas the maize gene has only
two introns that are at identical positions as two of the Arabi-
dopsis introns. Common motifs of these CKX genes are putative
N-terminal signal peptides, which indicate for most of the
corresponding proteins transport to the secretory pathway, and
a 70-aa-long FAD-binding domain in the N-terminal region
(data not shown; ref. 17).
Transgenic Plants Show Increased Cytokinin Oxidase Activity. We
cloned the AtCKX1, AtCKX2, AtCKX3, and AtCKX4 genes,
positioned them under the control of a constitutive 35S pro-
moter, transformed tobacco plants individually with these genes,
and selected overexpressing transgenic clones by Northern blot
analysis (Fig. 1A). The leaves of expressing transgenic tobacco
lines showed a 2.6-fold to 10.4-fold increase in cytokinin oxidase
activity (Fig. 1B). Likewise, cells of Saccharomyces cerevisiae
expressing AtCKX2 showed a higher cytokinin oxidase activity
than the control strain (Fig. 1C). The majority of the enzyme
activity accumulated in the yeast culture medium (Fig. 1C). This
result, and similar observations in fission yeast and Physcomi-
trella patens cells expressing the cytokinin oxidase of maize (9,
10), indicates that the cytokinin degradation pathway might be,
at least partially, located extracellularly. The apparent Km values
for the cytokinin oxidases analyzed in this study are in the range
of 0.3 to 9.5 M with iP as a substrate. This is similar to or even
lower than the maize enzyme, which has an apparent Km of 19.2
M in kernels (18) and a native Km of 2.8 M (17) with iP as a
substrate. These results demonstrate that the proteins encoded
by these four AtCKX genes do indeed have cytokinin oxidase
activity and could be used as a tool to study the relevance of
cytokinins during the whole life cycle of higher plants.
Transgenic Plants Have a Reduced Cytokinin Content. The endoge-
nous concentrations of different cytokinin metabolites was
significantly reduced in AtCKX1- and AtCKX2-expressing trans-
genic seedlings. The total content of iP and Z metabolites in
10488 www.pnas.orgcgidoi10.1073pnas.171304098
Fig. 1. AtCKX gene expression and enzyme activity in transgenic tobacco
plants and yeast. (A) Northern blots (50 g total RNA) of individual transfor-
mants were probed with gene-specific probes that covered the whole
genomic sequences. Only clones with detectable AtCKX transcripts showed a
phenotype, and no cross-hybridization with untransformed tobacco (WT) was
detected. 25S, hybridization with 25S rRNA as control for loading. (B) Cytoki-
nin oxidase activity in leaves of tobacco plants. Specific activity in extracts of wild
type (100%) was 8 0.9 pmol adenine mg protein1 h1. Bars show SD; n
3. (C) Cytokinin oxidase activity in yeast cells and medium. Specific activity of the
control strain (100%) was 1.16 nmol adenine mg protein1 h1.
individual transgenic clones was between 31% and 63% that of
wild type (Table 1). Among the 16 different cytokinin metab-
olites that were measured, the greatest change occurred in the
iP-type cytokinins in AtCKX2 overexpressers. Smaller alter-
ations were noted for Z-type cytokinins, which could be due to
different accessibility of the substrate or a lower substrate
specificity of the protein. Interestingly, the cytokinin reserve
pool of O-glucosides was also lowered in the transgenics (Table
1). The concentrations of N-glucosides and dihydrozeatin-type
cytokinins were very low in wild-type plants and were not, or only
marginally, altered in transgenic seedlings (data not shown). It
is noteworthy that the overall decrease in the content of iP-type
cytokinins is more pronounced in AtCKX2-expressing plants
than in AtCKX1 transgenics, which show a stronger phenotype in
the shoot. The changes in concentration of Z and ZR were
similar in both cases. It is not known which cytokinin metabolite
is relevant for the different traits that were analyzed in
the transgenic plants. It may be that the different cytokinin
forms have differing roles to play in the various developmen-
tal processes.
Transgenic Plants Have an Altered Phenotype. AtCKX gene overex-
pression caused striking developmental alterations in the plant
shoot and root system. The alterations were very similar, but
not identical, for the different genes. AtCKX1 and AtCKX3
overexpressers were alike as were AtCKX2 and AtCKX4 trans-
genics. Generally, the two former showed higher expression of
the traits, particularly in the shoot. Because of these similar-
ities, the phenotypic changes of two independent clones ex-
pressing AtCKX1 or AtCKX2 were analyzed in greater detail.
Most data shown refer to the stronger phenotype of the
AtCKX1 transgenics.
The most noticeable changes in the shoot were a severely
retarded development with shorter internodes leading to a
Werner et al.
Table 1. Cytokinin content of AtCKX transgenic plants
Line Wild type AtCKX1-2 AtCKX1-28 AtCKX2-38 AtCKX2-40

Cytokinin % of % of % of % of
metabolite Concentration Concentration WT Concentration WT Concentration WT Concentration WT

iP 5.90 1.80 4.76 0.82 81 4.94 2.62 84 1.82 0.44 31 2.85 0.62 48
iPR 2.36 0.74 1.53 0.14 65 0.75 0.27 32 0.55 0.39 23 0.89 0.07 38
iPRP 3.32 0.73 0.87 0.26 28 1.12 0.13 34 0.80 0.48 24 1.68 0.45 51
Z 0.24 0.06 0.17 0.02 71 0.22 0.03 92 0.21 0.06 88 0.22 0.02 92
ZR 0.60 0.13 0.32 0.12 53 0.34 0.03 57 0.34 0.15 57 0.32 0.05 53
ZRP 0.39 0.17 0.42 0.11 107 0.28 0.15 72 0.06 0.01 15 0.17 0.06 44
ZOG 0.46 0.20 0.32 0.09 70 0.26 0.13 57 0.20 0.07 43 0.12 0.02 26
ZROG 0.48 0.17 0.30 0.06 63 0.47 0.02 98 0.23 0.05 48 0.30 0.13 63

Total 13.75 8.69 63 8.38 61 4.21 31 6.55 48

Three independently pooled samples of approximately 100 2-week-old seedlings (2.5 gsample) were analyzed for each clone. Concentrations are in pmol
g fresh weight1. iP, N6-(2isopentenyl)adenine; iPR, N6-(2isopentenyl)adenine riboside; iPRP, N6-(2isopentenyl)adenine riboside 5-monophosphate; ZR,
zeatin riboside; ZRP, zeatin riboside 5-monophosphate; ZOG, zeatin O-glucoside; ZROG, zeatin riboside O-glucoside.

dwarfed growth habit, the formation of lanceolate epinastic
leaves, and the formation of a reduced number of flowers (Fig.
2 A and B). The time between the initiation of new leaves
(plastochrone) at the borders of the shoot meristem was on
average 2.6 0.1 days in wild type and 4.4 0.1 days in AtCKX1
transgenics (Fig. 2C). The surface area of leaves formed by the
transgenics during a defined time period was 15% that of wild
type (Fig. 2D). The width-to-length ratio of mature leaves was
lowered from 1:2 in wild type to 1:3 in AtCKX1 transgenics. The
vasculature of AtCKX1 transgenic leaves was less developed, the
spacing between veins was larger, and the veins were flat and not
raised as in wild type. In contrast to wild-type leaves, leaf
parenchyma cells continued to expand in transgenic clones in the
transverse direction, resulting in thicker and rigid old leaves. A
prominent difference was also noted for progression of leaf
senescence. In tobacco, leaf senescence starts in the most basal
leaves and leads to a uniform reduction of leaf pigment content.
In contrast, aging leaves of AtCKX1 transgenic plants developed
chlorotic intercostal regions but retained chlorophyll along the
leaf veins (Fig. 2E). Leaf aging was similar in AtCKX2-expressing
plants, but chlorosis was less pronounced. Transgenic plants
started to flower up to 3 months later than wild-type plants (Fig.
2C) and produced only 510 normal-sized flowers compared
with 100 flowers in the wild types. The final leaf number at the
onset of flowering was similar in wild type and the transgenic
clones, supporting the notion that leaf number is a determinant
for flower induction in day-neutral tobacco (Fig. 2C). Lateral
buds in the leaf axils of transgenic plants developed two to three
tiny leaves early during vegetative development, in contrast to
lateral buds of wild type, which remained completely inhibited.
This indicates incomplete apical dominance in the transgenic
plants.
In contrast to the inhibited shoot development of AtCKX
transgenic tobacco, their root growth was enhanced (Fig. 3 A and

PLANT BIOLOGY

Fig. 2. Shoot phenotype of AtCKX1-expressing tobacco plants. (A) Top view of 6-week-old plants. (B) Tobacco plants at the flowering stage. (C) Kinetics of stem
elongation. Arrows mark the onset of flowering. Age of plants (days after germination) and leaf number at that stage are indicated above the arrows. Bars
indicate SD; n 12. (D) Number of leaves (n 12) formed between day 68 and day 100 after germination and final surface area of these leaves (100% of wild
type is 3646 144 cm2; n 3). (E) Comparison of leaf size and senescence. Leaves were from nodes number 4, 9, 12, 16, and 20 from the top (from left to right).

Werner et al. PNAS August 28, 2001 vol. 98 no. 18 10489

Fig. 3. Root phenotype of AtCKX-expressing transgenic tobacco plants. (A) Seedlings 17 days after germination. (B) Root system of soil-grown plants at the
flowering stage. (C) Root length, number of lateral roots (LR), and adventitious roots (AR) on day 10 after germination. (D) Dose-response curve of root growth
inhibition by exogenous cytokinin. Seeds were sown on MS medium containing 3% sucrose and the indicated concentration of iPR. The length of primary roots
was determined after 10 days of cultivation in the dark on vertically positioned plates. Bars indicate SD; n 30.
B). Elongation of the primary root was more rapid, primordia of
lateral roots were noted closer to the root apex than in wild-type
plants, and the number of lateral branches, as well as adventi-
tious roots, increased (Fig. 3C). Enhanced root growth led to a
60% increase in root dry weight in transgenic plants grown in
hydroponic solution (data not shown). These results suggest that
cytokinins are involved in controlling both root growth rate and
the generation of new root meristems. The dose-response curve
of root growth inhibition by exogenous cytokinin showed the
transgenic roots to have cytokinin resistance (Fig. 3D). Inter-
estingly, the resistance of AtCKX1 transgenics to iPR was less
marked than for AtCKX2, which is consistent with the smaller
changes in iP-type cytokinins in the latter (Table 1).
Histology of the Shoot Meristem, Shoot Organs, and Root Meristems.
A decreased or increased organ growth rate as a consequence of
a reduced cytokinin content could be due to a changed cell
division rate in the meristematic regions, a different population
size of dividing cells, or altered cell growth. In the AtCKX
transgenics, the final length of cells in the stem was not reduced,
and the final length of root cells was slightly decreased (149.7
31.7 M in clone AtCKX1-50 versus 167.0 32.0 M in wild
type; n 100), indicating that differences in cell growth did not
contribute to, or even partially compensate for, altered growth
of stem and roots. However, microscopic inspection of the shoot
apical meristem (SAM), leaf, and the root meristem revealed
that the morphological changes described above were reflected
in distinct changes in cell number and rate of cell formation in
the AtCKX transgenics.
The SAM of AtCKX1 transgenic plants was smaller than in
wild-type plants and fewer cells occupied the space between the
central zone and the peripheral zone of lateral organ formation,
but the cells were of the same size and no obvious changes of the
differentiation pattern occurred (Fig. 4A). Also, the overall
tissue pattern of leaves in cytokinin oxidase overexpressers was
unchanged. However, the sizes of both phloem and xylem were
significantly reduced (Fig. 4B). In contrast, the average cell size
of leaf parenchyma and epidermal cells was increased 4- to 5-fold
(Fig. 4 C and D). New cells of AtCKX1 transgenic leaves are
formed at 34% of the rate of wild-type leaves, and final leaf cell
number is estimated to be in the range of 56% that of wild type.
Similar but less pronounced changes occurred in the shoot of
AtCKX2-expressing plants (data not shown). In contrast to
leaves, neither cell size nor cell form of floral organs was altered
in the transgenic lines. Also, seed weight was similar in wild type
and AtCKX1 and AtCKX2 transgenic plants (data not shown).
The cell population in root meristems in the AtCKX1 and
AtCKX2 transgenic plants was enlarged approximately 4-fold,
10490 www.pnas.orgcgidoi10.1073pnas.171304098
and the cell numbers in both the central and lateral columnella
were increased (Fig. 4 E and F). Final root diameter was
increased by 60% due to the increased diameter of all root cell
types and an increased number of cells in each cell file. The radial
root pattern was identical in wild type and transgenics, with the
exception that frequently a fourth layer of cortex cells was noted
in transgenic roots (Fig. 4G).
Discussion
This analysis of the consequences of reduced endogenous cyto-
kinin content strongly indicates in which plant processes cyto-
kinins are limiting and might, therefore, have a regulatory
function. The slowed formation of new cells in the SAM, as well
as of leaf primordia, and the reduced size of the SAM indicates
that cytokinins have a dual function in the control of SAM
proliferation. They are required to maintain the cell division
cycle but might also be involved in promoting the transition from
undifferentiated stem cells to differentiation. Earlier work has
shown that in unorganized growing cells, cytokinins induce the
formation of shoot meristems, demonstrating that they have a
function beyond maintaining the cell cycle (2). Known coordi-
nating factors of cell proliferation and differentiation in the
SAM are transmembrane receptor proteins (e.g., CLV1) and
transcription factors of the homeodomain class (e.g., WUS,
STM, KNAT1), which interact in regulatory loops (19). Recent
data indicate that a reciprocal interaction between cytokinins
and some of these transcription factors exists (2022). A role for
cytokinins in the regulation of SAM differentiation could be
realized through local gradients of the hormone or differences
in the distribution of different cytokinin metabolites. This might
alter effector gene expression quantitatively, which could in turn
influence cellular fate. Developmental changes in the concen-
tration and localization of different cytokinin metabolites have
been reported for the SAM of tobacco (23). The reduced activity
of the SAM could also be the cause of the incomplete apical
dominance, which was noted in transgenic plants, as the amount
of auxin produced for the maintenance of apical dominance
might be lowered.
The slowed formation of leaf cells and their reduced number
indicates an absolute requirement for cytokinins during leaf
formation, both to drive the cell division cycle at normal speed
and to obtain the required number of divisions for a normal leaf
size. That cytokinins function as a regulatory factor in leaf cell
formation is supported by the fact that transgenic Arabidopsis
plants with an enhanced cytokinin content produced more leaf
cells than control plants (20). Moreover, cytokinins appear to
restrict leaf cell size as the cells of transgenic leaves are larger
than in control plants. Alternatively, a compensatory mechanism
Werner et al.
Fig. 4. Histology of shoot meristems, leaves, and root meristems. (A) Longitudinal median section through the vegetative SAM. P, leaf primordia. (B) Vascular
tissue in second order veins of leaves. X, xylem, PH, a phloem bundle. (C) Cross sections of fully developed leaves. (D) Scanning electron microscopy of the upper
leaf epidermis. (E) Root apices stained with 4,6-diamidino-2-phenylindole. RM, root meristem. (F) Longitudinal median sections of root meristems 10 days after
germination. RC, root cap; PM, promeristem. (G) Transverse root sections 10 mm from the apex. E, epidermis, C1C4, cortical cell layer; X, xylem; PH, phloem.
The material for the analysis of the SAM and the mature fully expanded leaves was from 38- and 100-day-old plants (clone AtCKX1-50), respectively, which were
cultivated in a green house. Root analysis was performed with primary roots of seedlings 10 days after germination. Bars, 100 m.

may be activated in transgenic plants to reach a genetically
determined organ size, as has been reported for plants express-
ing a dominant-negative form of cdc2 (24). In either case, the leaf
phenotype of AtCKX overexpressers supports the view that cell
cells rather than increased cell growth. In the presence of
lowered cytokinin content, root meristem cells have a prolonged
meristematic phase and eventually undergo additional rounds of
mitosis before they leave the meristem and start to elongate. We
PLANT BIOLOGY

proliferation and growth in tobacco leaves are not coupled.
Interestingly, the flower phenotype of the transgenic plants
was unaltered. This suggests that the role of cytokinins in the
regulation of development of reproductive organs might be less
important than it is during the vegetative phase. It may be that
once the plant has entered the reproductive cycle, a more
stringent mechanism operates in the meristem to ensure the
proper course of the developmental program.
Contrasting with the promotive role in the SAM, cytokinins
have a negative regulatory function in root growth. The in-
creased cell number in the transgenic root meristems and the
slightly reduced final cell length in transgenic roots indicate that
the enhanced root growth is because of an enhanced cycling of
conclude that the activity of the initials andor the exit of cells
from the root meristem is regulated by a mechanism that is
sensitive to cytokinins.
Taken together, the investigation of cytokinin-deficient plants
has shown that the influence of cytokinins on morphogenesis is
primarily achieved through cell cycle regulation. Multiple func-
tions and several molecular targets of cytokinins during different
phases of the cell cycle are known. The hormone is required for
S-phase entry in leaf mesophyll protoplasts and tobacco pith
explants, and S-phase progression is accelerated in the presence
of cytokinins (25, 26). Several cell cycle genes are regulated by
cytokinins, including cdc2, CycD3, and different B-type cyclins
(2729). There is evidence that regulatory genes of the cell cycle

Werner et al. PNAS August 28, 2001 vol. 98 no. 18 10491

are expressed in a tissue-specific fashion and that cytokinin
effects on the cell cycle vary between different cell types (30, 31).
Distinct expression patterns of cytokinin targets could be a
reason for the opposite effects seen in shoot and root meristems.
The enhanced organ growth in plants overexpressing D- and
B-type cyclins (32, 33) is consistent with the hypothesis that
cytokinins act through the regulation of cell cycle progression.
What could the role of CKX proteins during plant growth and
development be? A likely role is the degradation of cytokinins
that accumulate transiently during the G2M transition of cy-
cling cells to a level that is several orders of magnitude higher
than during the other cell cycle phases (34). It is not known how
these cytokinin levels are rapidly readjusted to normal levels.
CKX enzymes could have a role in recycling this cell division-
derived cytokinin. The expression of AtCKX genes in regions of
active growth is consistent with the proposed function in cycling
cells (unpublished results). Additional roles for the enzymes
could be the maintenance of an optimal level of cytokinins for
growth andor resetting a cytokinin signaling system to a basal
level.
To summarize, in this work we have obtained proof of the
function of four cytokinin oxidases from A. thaliana and used
these genes as tools to generate plants with a reduced cytokinin
content. The data lend support to the auxincytokinin hypoth-
1. Miller, C. O., Skoog, F. S., Okomura, M. H., Von Saltza, H. & Strong, F. M.
(1955) J. Am. Chem. Soc. 77, 13921393.
2. Skoog, F. & Miller, C. O. (1957) Symp. Soc. Exp. Biol. 11, 118131.
3. Letham, S. (1994) in Cytokinins: Chemistry, Activity and Function, eds. Mok,
D. W. S. & Mok, M. C. (CRC, Boca Raton, FL), pp. 5780.
4. Mok, M. C. (1994) in Cytokinins: Chemistry, Activity and Function, eds. Mok,
D. W. S. & Mok, M. C. (CRC, Boca Raton, FL), pp. 155166.
5. Klee, H. J. & Lanahan, M. B. (1995) in Plant Hormones: Physiology, Biochem-
istry and Molecular Biology, ed. Davies, P. J. (Kluwer, Dordrecht, the Nether-
lands), pp. 340353.
6. Schmu lling, T., Rupp, H. M., Frank, M. & Schafer, S. (1999) in Advances in
Regulation of Plant Growth and Development, eds. Strnad, M., Pec P. & Beck,
E. (Peres, Prague), pp. 8596.
7. Hooykaas, P. J. J., Hall, M. A. & Libbenga, K. R., eds. (1999) Biochemistry and
Molecular Biology of Plant Hormones (Elsevier, Amsterdam).
8. Armstrong, D. J. (1994) in Cytokinins: Chemistry, Activity and Function, eds.
Mok, D. W. S. & Mok, M. C. (CRC, Boca Raton, FL), pp. 139154.
9. Morris, R. O., Bilyeu, K. D., Laskey, J. G. & Cheikh, N. N. (1999) Biochem.
Biophys. Res. Commun. 255, 328333.
10. Houba-Herin, N., Pethe, C., dAlayer, J. & Laloue, M. (1999) Plant J. 17,
615626.
11. Gatz, C., Frohberg, C. & Wendenburg, R. (1992) Plant J. 2, 397404.
12. Horsch, R. B., Fry, J. E., Hoffman, N., Wallroth, M., Eichholtz, D., Rogers, S.
& Fraley, R. T. (1985) Science 227, 11291132.
13. Faiss, M., Zalubilova, J., Strnad, M. & Schmu lling, T. (1997) Plant J. 12,
401415.
14. Schoof, H., Lenhard, M., Haecker, A., Mayer, K. F. M., Ju rgens, G. & Laux,
T. (2000) Cell 100, 635644.
15. Neinhuis, C. & Edelmann, H. G. (1996) J. Microsc. (Oxford) 184, 1416.
16. Motyka, V., Faiss, M., Strnad, M., Kamnek, M. & Schmu lling, T. (1996) Plant
Physiol. 112, 10351043.
17. Bilyeu, K. D., Cole, J. L., Laskey, J. G., Riekhof, W. R., Esparza, T. J., Kramer,
M. D. & Morris, R. O. (2001) Plant Physiol. 125, 378386.
18. McGaw, B. A. & Horgan, R. (1983) Planta 159, 3037.
19. Fletcher, J. C. & Meyerowitz, E. M. (2000) Curr. Opin. Plant Biol. 3, 2330.
10492 www.pnas.orgcgidoi10.1073pnas.171304098
esis for quantitative growth parameters and organ ratio in plants
(2). However, we note that some, but not all, phenotypic changes
can be explained by an altered ratio of these two hormones, as
several aspects of auxin-overproducing plants are distinct from
plants with a reduced cytokinin content (35, 36). This indicates
that the auxincytokinin balance determines only a subset of
morphogenetic parameters. Natural cytokinin levels are inhib-
itory to the development of a maximal root system, and fine
adjustment of cytokinin levels is needed to achieve the optimal
growth of shoots (20, 37). Evidently, the targeted manipulation
of CKX gene expression can be an important and novel tool to
modulate growth characteristics and yield parameters of crop
plants.
We dedicate this article to C. O. Miller and the late F. S. Skoog, who
discovered cytokinins almost 50 years ago and coined the auxin
cytokinin hypothesis of plant growth. We are indebted to K. Lemcke for
initial help with gene cloning and sequence analyses, M. Riefler for
support with structural gene analysis, M. Lenhard and Y. Stierhof for
advise for microscopic analyses, and M. Kamnek for helpful comments
on the manuscript. We thank H. Martnkova and V. Lacmanova for
excellent technical assistance and M. J. Beech and C. Scott-Taggart for
proofreading. We acknowledge financial support of the Deutsche For-
schungsgemeinschaft (Schm 81413-1), the Volkswagen-Stiftung (I
72076), the Grant Agency of the Czech Republic (522001346), and the
Czech Ministry of Education (MSM 153100008).
20. Rupp, H. M., Frank, M., Werner, T., Strnad, M. & Schmu lling, T. (1999) Plant
J. 18, 357363.
21. Tamaoki, M., Kusaba, S., Kano-Murakami, Y. & Matsuoka, M. (1997) Plant
Cell Physiol. 38, 917927.
22. Ori, N., Juarez, M. T., Jackson, D., Yamaguchi, J., Banowetz, G. M. & Hake,
S. (1999) Plant Cell 11, 10731080.
23. Dewitte, W., Chiapetta, A., Azmi, A., Witters, E., Strnad, M., Rembur, J., Noin,
M. & Chriqui, D. (1999) Plant Physiol. 119, 111121.
24. Hemerly, A., de Almeida Engler, J., Bergounioux, C., Van Montagu, M.,
Engler, G., Inze, D. & Ferreira, P. (1995) EMBO J. 14, 39253936.
25. Cooke, R. & Meyer, Y. (1981) Planta 152, 17.
26. Jacqmard, A., Houssa, C. & Bernier, G. (1994) in Cytokinins: Chemistry, Activity
and Function, eds. Mok, D. W. S. & Mok, M. C. (CRC, Boca Raton, FL), pp.
197215.
27. Hemerly, A. S., Ferreira, P., de Almeida Engler, J., Van Montagu, M., Engler,
G. & Inze, D. (1993) Plant Cell 5, 17111723.
28. Riou-Khamlichi, C., Huntley, R., Jacqmard, A. & Murray, J. A. H. (1999)
Science 283, 15411544.
29. Jelenska, J., Deckert, J., Kondorosi, E. & Legocki, A. B. (2000) Plant Sci. 150,
2939.
30. Gaudin, V., Lunners, P. H., Fobert, R. P. R., Towers, M., Riou-Khamlichi, C.,
Murray, J. A. H., Coen, E. & Doonan, J. H. (2000) Plant Physiol. 122,
11371148.
31. Coenen, C. & Lomax, T. L. (1988) Plant Physiol. 117, 6372.
32. Cockkroft, C. E., den Boer, B. G. W., Healy, J. M. S. & Murray, J. A. H. (2000)
Nature (London) 405, 575579.
33. Doerner, P., Jorgensen, J. E., You, R., Steppuhn, J. & Lamb, C. (1996) Nature
(London) 380, 520523.
34. Redig, P., Shaul, O., Inze, D., Van Montagu, M. & Van Onckelen, H. (1996)
FEBS Lett. 391, 175180.
35. Klee, H. J., Horsch, R. B., Hinchee, M. A., Hein, M. B. & Hoffmann, N. L.
(1987) Genes Dev. 1, 8696.
36. Sitbon, F., Hennion, S., Sundberg, B., Little, C. H. A., Olsson, O. & Sandberg,
G. (1992) Plant Physiol. 99, 10621069.
37. Gan, S. & Amasino, R. M. (1995) Science 270, 19861988.
Werner et al.